We consider the recent formulation of the Algorithmic Lov\'asz Local Lemma [9,2] for finding objects that avoid "bad features", or "flaws". It extends the Moser-Tardos resampling algorithm  to more general discrete spaces. At each step the method picks a flaw present in the current state and "resamples" it using a "resampling oracle" provided by the user. However, it is less flexible than the Moser-Tardos method since [9,2] require a specific flaw selection rule, whereas  allows an arbitrary rule (and thus can potentially be implemented more efficiently). We formulate a new "commutativity" condition, and prove that it is sufficient for an arbitrary rule to work. It also enables an efficient parallelization under an additional assumption. We then show that existing resampling oracles for perfect matchings and permutations do satisfy this condition. Finally, we generalize the precondition in  (in the case of symmetric potential causality graphs). This unifies special cases that previously were treated separately.